How Does Speed Effect Efficiency/Range?

[Jason McNeil]

In this set of tests, I wanted to understand the relationship between maintaining a steady cruising speed & the motor's avg power consumption over two laps (800M) of a running track.

The physics of power consumption on a conventional momentum balanced device (e.g. a bicycle) are well understand, but not so with the single-wheel self-balanced systems. There is some debate over how much, or even if, there is an energy penalty to keep the rider upright, or if it conserved in forward momentum.

http://www.gribble.org/cycling/power_v_speed.html (watts) = (1-(Loss_dt/100))^-1 · (F_gravity + F_rolling + F_drag) · V (m/s)

Another question that needed answering was, 'does a more powerful motor have an impact on efficiency & if so, by what degree?'

Methods: 

1. The two eWheels used for the testing were the Ninebot E, 240Wh battery & the new King Song 800W with the 680Wh battery pack.
2. An inline power meter capturing Volts, current, temperature between the battery pack & the control board, recording at a rate of 50Hz  
3. Two laps around the track maintaining speed intervals of 15kph, 18kph (for the Ninebot cannot maintain 20kph), 20kph, 25kph, & 28kph
4. The average power consumption is taken across a minimum of 2 minutes, removing the start (acceleration) & finishing outliers

Results:

1. At the same speeds, 15kph for instance, both the Ninebot & King Song are practically identical at 12Wh/km
2. There is no significant diminishing rate in efficiency as the motor is operated at the software defined top-speed of 28kph that can not be accounted for by wind resistance

 

[Gimlet]

This subject has been addressed on the Tesla forums where the conclusion on range-Wh/km with the lower powered models and the higher powered models isn't that different when driven in the same manner, the lower powered ones gained marginally but mainly due to the lower weight of batteries fitted.

This is very different now though, with the latest "D" models with two motors that can be individually turned off when not required to save power.

[Kevin Lee]

May we have a look at your connection setup? What equipment did you capture the V/A/T with and with what capturing software? I just want to verify the measuring methodologies and the voltage and current drops across the bridging cables. What cable gauge did you use? This one maybe? http://www.eagletreesystems.com/index.php?route=product/product&path=62&product_id=54

I think we first have to verify the accuracy of the eLogger, then do a bit more logging under different conditions like downhill, flat, and uphill under a few acceleration methods, then we will be able to tell the power consumption of an 800W motor against a 500W. However, this will not give an accurate figure as Ninebot's controller board may not be aggressive as KingSong's.

[Jason McNeil]

@Kevin, wrote about it here... http://forum.electricunicycle.org/topic/851-data-capturing-performance-guide-for-ewheels/#comment-8512

If you look around the forums, I've been using the eLogger for over a month, the testing method is good enough for what I was trying to capture, you're more than welcome to do your own performance analysis for other scenarios    

this will not give an accurate figure as Ninebot's controller board may not be aggressive as KingSong's.

Don't quite follow this remark? What are you referring to? The data logging is occurring between the battery and controller, it's a precise way of recording power usage independently of any App or vendor. 

[robca]

May we have a look at your connection setup? What equipment did you capture the V/A/T with and with what capturing software? I just want to verify the measuring methodologies and the voltage and current drops across the bridging cables. What cable gauge did you use? This one maybe? http://www.eagletreesystems.com/index.php?route=product/product&path=62&product_id=54

I think we first have to verify the accuracy of the eLogger, then do a bit more logging under different conditions like downhill, flat, and uphill under a few acceleration methods, then we will be able to tell the power consumption of an 800W motor against a 500W. However, this will not give an accurate figure as Ninebot's controller board may not be aggressive as KingSong's.

Whatever bias/fault/poor calibration the elogger might have (and I'm not saying it does, just that it might), the results are consistent from run to run (@Jason McNeil has shared many logging sessions in the past). As such, the results are comparable (even if we can't say if the measured W are truly correct, we can say that the difference between lower speeds and higher speeds is a certain %. We can also compare the Ninebot to KS, since they have very similarly recorded power curves...)

I agree with you that logging sessions on flat terrain might yield very different results from runs in uphill, downhill and more importantly mixed up/down terrain (I suspect, but have no data to prove it, that some of the extra high regen numbers we have seen in some of the past runs will not result in real energy stored in the battery: Li-Ion batteries can't be charged at much more than 1C, so when I see big regen peaks, I suspect that the BMS simply finds a way to "bleed" some energy away... in this case, a more powerful motor might in theory be similar to a less powerful one when you measure the current balance in theory, but a less powerful motor might end up being more efficient, by using less power uphill, and regenerating most of it downhill, while a very powerful motor might use more power uphill and not recover all of it in the battery when going downhill... unfortunately, that would be very hard to measure and would require measuring the current going into each cell)

[zlymex]

I installed a precision current sensor(LEM ITS 25-NP)between the battery and controller of my MSuper V3 to obtain the current signal. The voltage signal is from the charge port. I use GRAPHTEC GL220 Data Logger to log the current and voltage at 100ms interval.

I rode in a section of a straight road back and forth for each speed, and use Wheellog to monitor the speed while I was ridding.

The result is very similar at high speed as OP, but I achieve below 10Wh per km at lower speed probably because my weight is on the lower side and I inflated my tire a little bit higher(3.0 bar)

I also tested at 20km/h when I was carrying an 25kg backpack to simulate a 100kg rider.The consumption went up by 1.3Wh per km  
I also tested at 20km/h when the tire pressure lowered to 2.0 bar, and the consumption went up by 0.9Wh per km